Do Changes in Visceral Sensory Function Determine the Development of Dyspepsia During Treatment With Aspirin?

Transcription

1 GASTROENTEROLOGY 2002;123: Do Changes in Visceral Sensory Function Determine the Development of Dyspepsia During Treatment With Aspirin? GERALD HOLTMANN,* JUERGEN GSCHOSSMANN,* LYDIA BUENGER,* GUIDO GERKEN,* and NICHOLAS J. TALLEY *Division of Gastroenterology and Hepatology, University of Essen, Essen, Germany; and the Department of Medicine, University of Sydney, Nepean Hospital, Sydney, Australia Background & Aims: We hypothesized that the development of dyspeptic symptoms during treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) would be linked to alterations in gastric mechanosensory function and gastric emptying. Methods: In the first study, gastric mechanosensory thresholds (barostat technique) and gastric emptying ( 13 C-octanoic breath test) were measured and endoscopy was performed at entry and after 5 days of treatment with aspirin (500 mg 3 times daily) in 8 patients with functional dyspepsia (initially without symptoms) and 8 healthy controls. In a second, doubleblind, placebo-controlled, cross-over study, 6 new patients with functional dyspepsia and 6 controls were started with either placebo or aspirin for 5 days. Sensory thresholds were tested after the fifth day of aspirin or placebo treatment. Abdominal symptoms were assessed daily. Results: In the first study, 6 of 8 patients and 3 of 8 controls, and in the second trial 6 of 6 patients and 1 of 6 healthy subjects, developed dyspepsia on aspirin (P < patients vs. healthy subjects). No symptoms occurred during placebo treatment. Lanza scores were not associated with symptoms. After aspirin, sensory thresholds increased in both studies in subjects without development of symptoms (by 25.9% 7.9%, and 31.0% 4.1%, respectively, all P < 0.05), whereas there was no significant increase in subjects who developed symptoms ( 11.2% 5.3% and 3.4% 13.4%, all P > 0.4). Neither thresholds nor symptoms were linked with the severity of mucosal damage, baseline gastric emptying (t1/2), or changes of gastric emptying (all P > 0.4). Conclusions: Failure to increase sensory thresholds during treatment with aspirin is associated with the development of dyspepsia. Acute treatment with traditional nonsteroidal antiinflammatory drugs (NSAIDs) is usually associated with the development of mucosal lesions and often dyspeptic symptoms. 1 It is well recognized that the extent and severity of erosions is not directly associated with an increased risk for dyspeptic symptoms or more severe symptoms. 2 Recent epidemiologic studies have supported an increased risk for the development of dyspeptic symptoms during treatment with NSAIDs. 3 Because large numbers of people are being started on aspirin for cardiovascular prophylaxis, aspirin-induced dyspepsia remains an important clinical issue. 4 Altered visceral mechanosensory function (hyperalgesia) is believed to play a key role in the development of symptoms in at least a subgroup of patients with otherwise unexplained (functional) abdominal complaints The mechanisms involved in the development of visceral hypersensitivity are as yet not completely determined, but sensitization by mucosal inflammation, impaired central down-regulation of visceral afferents, 11,12 or altered central processing 13 may be important. Recently, we have observed a failure of mechanosensory thresholds to increase after repeated gastric mechanical stimulation in patients with functional dyspepsia compared with healthy subjects. 14 These data are consistent with the concept that the activation of antinociceptive mechanisms in response to visceral stimulation is disturbed in at least one subgroup of patients with functional dyspepsia. In addition to alterations of visceral sensory function, disturbances of gastric emptying also may play a role in the development of symptoms in functional dyspepsia, 15 although this remains controversial. Because aspirin alters prostaglandin metabolism and prostaglandins may be involved in the regulation of gastrointestinal motility (e.g., prostaglandins administered orally can induce phase III complexes 16 ), acute changes in gastric emptying also may be relevant in the pathogenesis of aspirininduced dyspepsia. We hypothesized that the development of symptoms during treatment with aspirin would be linked to alterations of visceral sensory function or changes in gastric emptying. We aimed to test this hypothesis in currently Abbreviation used in this paper: MDP, minimal distending pressure by the American Gastroenterological Association /02/$35.00 doi: /gast

2 1452 HOLTMANN ET AL. GASTROENTEROLOGY Vol. 123, No. 5 asymptomatic patients with well-documented functional dyspepsia and healthy subjects. Because sensory testing studies are time consuming and technically challenging, we decided to follow a 2-step approach. After an initial study to test the hypothesis and to determine an appropriate sample size, a second study was conducted to further validate the results. Materials and Methods Study A: Single-Blind Study Patients and healthy subjects. Eight patients with functional dyspepsia (recurrent upper-abdominal pain or discomfort for more than 2 years, mean age 26.3 years, range years, 4 women) and 8 age- and gender-matched healthy volunteers were recruited for participation in this study by public advertisements. All patients had a history of symptoms for more than 5 years and all patients had overlapping symptoms of dysmotility (e.g., early satiety, fullness, nausea, symptoms aggravated by meals) and ulcer-like functional dyspepsia (e.g., pain localized in the upper abdomen, pain relieved by food). All subjects were required to be symptom free for more than 2 weeks before inclusion in the study and had to be off any medications affecting the gastrointestinal tract for at least this time period. Helicobacter pylori status was determined by using a 13 C urea breath test. All patients with functional dyspepsia had a normal physical examination and no alarm symptoms (e.g., no weight loss) and did not report irritable bowel symptoms such as altered bowel habits or pain associated with defecation. Laboratory testing (i.e., white and red blood count, erythrocyte sedimentation rate, fasting blood glucose level, liver function tests) was normal in all subjects. Protocol. After approval of the research protocol by the local Ethics Committee and informed consent, the subjects were studied on 2 different days. On days 1 and 7, visceral sensory thresholds and gastric emptying rates were determined. From days 2 6, all study subjects were treated in a single blinded fashion (i.e., patients and healthy subjects were unaware whether active drug or placebo were administered) with acetyl salicylic acid (Aspirin, Bayer AG, Leverkusen, Germany) 500 mg 3 times daily. The subjects were told that they could receive placebo or active drug. Each day, 3 symptoms of dyspepsia, namely pain, nausea, and discomfort (i.e., an unpleasant sensation not reaching the level of pain) were assessed and intensity of symptoms was graded (no symptoms, mild, moderate, severe, very severe symptoms). If for any of these symptoms the study subject reported at least mild severity, this subject was considered to have dyspepsia. In 8 randomly selected subjects from the study population, a third measurement of abdominal symptoms was conducted after 5 days of treatment with placebo tablets (3 tablets per day, identical in appearance to aspirin) that were administered in a single blinded fashion. On day 6, sensory thresholds and gastric emptying were measured again. The follow-up studies were conducted to determine whether repeated barostat or Figure 1. Schematic representation of the study design of study A. Eight patients with functional dyspepsia (FD) and 8 healthy subjects were treated for 5 days with aspirin (500 mg 3 times daily), and sensory function and gastric emptying were assessed before and after aspirin treatment. In addition, in 8 randomly selected subjects (4 with functional dyspepsia and 4 healthy subjects), sensory function and gastric emptying were assessed after another 5 days of placebo treatment. gastric emptying testing per se induced systematic alterations of gastric emptying or sensory function. The study subjects were all informed at entry that they could be treated either with aspirin or placebo. The study protocol is depicted in Figure 1. Upper gastrointestinal endoscopy. An upper-abdominal endoscopy was performed (by using an Olympus Q140 endoscope; Olympus, Hamburg, Germany) at baseline and between 2 6 hours after the second and, if applicable, after the third sensory test, to assess mucosal lesions. Photographs of the mucosa in the stomach and duodenum were taken applying a standard protocol, with at least one image taken from the fundic area, the corpus, the antrum, and the duodenal bulb. After completing the study, the extent of lesions was quantitated based on the pictures by using the Lanza score 17 (no visible lesions, score 0; 1 hemorrhage or erosion, score 1; 2 10 hemorrhages or erosions, score 2; hemorrhages or erosions, score 3; 25 hemorrhages or erosions, score 4) by 2 investigators independently (G.H. and L.B.). These investigators were blinded with regard to the treatment period and symptomatic status of the subjects. In case of discordant grading, the findings were discussed and a common grading was reached. Assessment of sensory thresholds. For the assessment of visceral sensory thresholds, the subjects swallowed an orogastric tube after an overnight fast ( 16 h after the last dose of medication). The barostat had an infinitely compliant bag attached to the tube, with a maximal capacity of 1.5 L, and the bag was positioned in the proximal part of the stomach. The subjects were all seated in an upright position. The barostat bag with a double-lumen tube was connected to the

3 November 2002 DYSPEPSIA DEVELOPMENT WITH ASPIRIN 1453 barostat device (Isobar 3 device, G & J Electronics, Ontario, Canada). After positioning the bag, the minimal distending pressure that allowed the recording of changes of intra-abdominal pressure associated with breathing was determined by increasing the pressure of the barostat bag in 1 mm Hg increments. To determine sensory thresholds, a distention procedure was performed. All subjects received standardized instructions. They were told that the balloon in the stomach would be inflated and deflated repeatedly and they were asked to indicate whether or not they actually perceived this distention stimulus by pressing the appropriate button on a keypad after a light signal. With ramp distention, the approximate threshold for first perception was determined. Thereafter, the distention procedure was started with 60% of the pressure of the expected sensory threshold. In random order, the distention pressure was either maintained or increased in 2 mm Hg increments and each pressure step was maintained for 45 seconds. At the end of the 45-second period, the light signal requested the subjects to press the appropriate button on the keypad. Thereafter, the distending pressure was lowered to 5 mm Hg and maintained at this level for 15 seconds. If the subjects had not perceived the preceding distention, the pressure was randomly either increased to the previous distention pressure or increased 2 mm Hg above this level. As soon as the subject reported perception of the stimulus, the pressure of the next distention was randomly maintained or lowered, and the subjects were again asked whether or not they perceived the distention. Subjects were kept unaware of the pressure level at each step. The threshold for first perception was defined as the distending pressure when on average the volunteer s response changed from no perception to distention perceived. These measurements of the threshold of first perception were repeated once within 90 minutes. The distentions were performed with a modified Isobar 3 device (G & J Electronics) by using the distender program. Previous experiments have shown excellent retest reliability for gastric sensory measurements repeated within 2 weeks in our laboratory (n 12, r 0.85, unpublished results). Gastric emptying test. The day before the barostat study and after a 12-hour fast (60 min after the last aspirin dose), the subjects ate a scrambled egg on 2 slices of bread with 20 g of butter. The egg was labeled with 100- g 13 C-octanoic acid. The test meal was eaten within 10 minutes and after completing the meal breath samples were taken in 10-minute intervals and for the remaining 3 hours at 15-minute intervals. 18 Bleeding time. Bleeding time was measured at baseline and on day 6. This was performed to measure the compliance of the subjects. Study B: Double-Blind Cross-Over Trial Patients and healthy subjects. After the results of study A, we aimed to confirm our findings by evaluating a new cohort of 12 subjects (6 with documented functional dyspepsia and 6 healthy subjects) who were also recruited by public Figure 2. Schematic representation of the study design of study B. After diagnostic work-up (including upper gastrointestinal endoscopy in patients with functional dyspepsia), patients and healthy subjects were randomized to treatment for 5 days either with aspirin (500 mg 3 times daily) or placebo (3 times daily). Thereafter, a barostat test was conducted and subjects were switched from aspirin to placebo or vice versa after a 36-hour washout period. advertisement after approval of the study protocol by the Institutional Ethical Board. These study subjects fulfilled the same inclusion and exclusion criteria as outlined for study A. The subjects in study B received aspirin or placebo in a randomized, double-blind, placebo-controlled, cross-over study (Figure 2). Both treatment arms were separated by a 36-hour washout period without medication. Abdominal symptoms were assessed at day 5 of aspirin treatment. Sensory thresholds were tested after 5 days of aspirin or placebo treatment. The same methodology was used for the sensory testing as described earlier. Statistical analyses. To assess the association between the extent of mucosal lesions and development of symptoms, 2 testing was performed. For the calculations of gastric wall tension, the balloon in the stomach was considered to be of spheric shape. The radius (R) of the bag was estimated from the volume of the sphere, and wall tension was calculated from the estimated bag radius and the intragastric pressure (P) by using Laplaces s law (T P R/2). 19 Analysis of variance for repeated measurements or nonparametric tests when appropriate were used to compare thresholds and gastric emptying in asymptomatic and symptomatic subjects as well as to determine the effects of the ramp distention in the different groups. All P values calculated were 2-tailed; an -level of 5% was considered significant. Statistical analyses were performed by using the Statistical Analysis System (SAS, Cary, NC). 20 Studies A and B were analyzed separately. Based on an a priori decision, the data on sensory thresholds were then pooled for a combined post hoc analysis. This was conducted to yield a more precise estimate of the effect of aspirin on sensory thresholds, and the association of alterations of sensory thresholds with the development of symptoms on aspirin treatment. Power calculation. Based on previous studies, 14,21 a sample size of 6 subjects per group would be sufficient to verify differences of sensory thresholds of more than 20% magnitude at the 5% level.

4 1454 HOLTMANN ET AL. GASTROENTEROLOGY Vol. 123, No. 5 Results Study A Symptoms on treatment with aspirin. Six of the 8 patients with functional dyspepsia compared with 3 of 8 healthy subjects on aspirin experienced dyspeptic symptoms. All patients with functional dyspepsia who became symptomatic reported upper-abdominal discomfort that was mild (n 2) or moderate (n 4). In healthy subjects, dyspepsia was mild in all but one patient who had moderate symptoms. One patient with functional dyspepsia also reported mild nausea as well as discomfort. Symptoms disappeared in all subjects within 48 hours after discontinuation of aspirin treatment. Mucosal lesions. All subjects had a normal endoscopy with no erosions at baseline. After 5 days of treatment with aspirin, all subjects were found to have mucosal lesions. The severity of the mucosal lesions in patients and healthy subjects as well as the severity of lesions in subjects with and without dyspeptic symptoms during treatment with aspirin are depicted in Figure 3. There were no differences between healthy subjects and patients with a history of functional dyspepsia, or between subjects with and without the development of symptoms during aspirin treatment (P 0.4). All lesions disappeared after 1 week in the subjects recruited for the placebo follow-up evaluation. Minimal distending pressure. The minimal distending pressure (MDP) and respective volumes that allowed recording of intra-abdominal pressure and volume changes associated with breathing were mm Hg and ml in healthy subjects, and mm Hg and ml in subjects with Figure 3. Study A. Severity of mucosal lesions after 5 days of treatment with aspirin in healthy subjects and patients with functional dyspepsia (upper panel) and subjects with and without development of symptoms during treatment with aspirin (lower panel). Figure 4. Study A. Thresholds for first perception before and after 5 days of aspirin treatment (FD, functional dyspepsia). The left panel depicts FD patients and healthy subjects without development of symptoms, the right panel indicates subjects with development of symptoms during treatment with aspirin. functional dyspepsia (all P 0.4). There were no statistically significant differences in MDP between patients and healthy subjects, or between subjects with and without symptoms during treatment with aspirin. Perception threshold (tracking). Mean thresholds significantly increased (P 0.05) in healthy subjects (from to mm Hg) whereas they tended to decrease (from to mm Hg, P 0.07) in patients with functional dyspepsia. Thus, the differences in perception thresholds between healthy subjects and patients with functional dyspepsia were significant after 5 days of treatment with aspirin (P 0.02), although they were not significantly different at the beginning of the study (P 0.2). Figure 4 depicts the sensory thresholds (pressure) before and after treatment with aspirin. The left panel depicts all subjects (healthy subjects and functional dyspepsia patients) without symptoms during treatment with aspirin, whereas the right panel shows the thresholds in all subjects reporting dyspeptic symptoms. Before aspirin treatment there were no significant differences between healthy subjects and patients with functional dyspepsia (P 0.4). In subjects who described no symptoms during treatment with aspirin, there was a significant increase of the sensory threshold (from to mm Hg, P 0.02), whereas there was no significant change in subjects reporting symptoms ( vs mm Hg, P 0.8). For the follow-up evaluation on placebo in study A, 4 healthy subjects and 4 patients with functional dyspepsia were randomly selected. Five came from the group without dyspeptic symptoms on aspirin (4 healthy subjects, 1 patient) and 3 from the group with dyspeptic symptoms during aspirin (all functional dyspepsia patients). During the placebo follow-up treatment period, sensory thresholds returned to baseline (without symptoms

5 November 2002 DYSPEPSIA DEVELOPMENT WITH ASPIRIN 1455 Table 1. Mean ( SEM) Gastric Emptying (t1/2) Time (min) Based on the 13 C-Octanoic-Breath Test Healthy subjects Functional dyspepsia patients Baseline 65 3 (n 8) (n 8) Aspirin 57 8 (n 8) (n 8) Placebo 53 5 (n 4) (n 4) No dyspepsia during aspirin challenge With dyspepsia during aspirin challenge Baseline 69 5 (n 7) (n 9) Aspirin 56 6 (n 7) (n 9) Placebo 51 4 (n 5) (n 3) NOTE. Upper part, healthy subjects vs. functional dyspepsia patients. Lower part, subjects without and with dyspepsia during aspirin challenge. There was a significant(p 0.05) decrease of the t1/2 after aspirin treatment compared with baseline. Differences between healthy subjects and patients and differences between subjects with and without dyspepsia before and during treatment with aspirin failed significance (P 0.2). mm Hg vs. with symptoms mm Hg) within 1 week (P 0.4 vs. baseline, Figure 4). Wall tension. Before aspirin treatment, gastric wall tension for perception was not significantly different between healthy subjects and patients with a history of functional dyspepsia (wall tension vs mm Hg/cm 2,F 0.8, P 0.8). During treatment with aspirin, wall tension was significantly lower in subjects with symptoms compared with subjects without symptoms (P 0.05). Gastric emptying. Gastric emptying before and after treatment with aspirin is shown in Table 1. During treatment with aspirin, there was a very small but statistically significant decrease of the gastric half-emptying time from 79 to 64 minutes (P 0.05). However, no significant alterations of gastric emptying were detected in subjects with and without symptoms on aspirin treatment (Table 1), or between healthy subjects and patients (P 0.4, Table 1). H. pylori status and sensory function or gastric emptying. Only 1 patient and 2 healthy subjects tested positive for H. pylori. H. pylori infected subjects were within the range of H. pylori negative subjects with regard to gastric emptying and sensory thresholds. Bleeding time. Bleeding time increased from a mean of seconds to seconds during treatment with aspirin (P 0.02). There were no significant differences in the bleeding times or changes of bleeding time during treatment with aspirin for patients or healthy subjects, or in subjects with and without symptoms during treatment. Study B Symptoms on treatment with aspirin. All 6 patients with functional dyspepsia and one of the healthy subjects developed dyspeptic symptoms during treatment with aspirin. Symptoms disappeared in all subjects within 48 hours after discontinuation of aspirin treatment. Minimal distending pressure. The MDP and respective volumes that allowed recording of intra-abdominal pressure and volume changes associated with breathing were mm Hg and ml in healthy subjects, and mm Hg and ml in subjects with functional dyspepsia (all P 0.4). There were no statistically significant differences in MDP between patients and healthy subjects, or between subjects with and without symptoms during treatment with aspirin. Perception threshold (tracking). Figure 5 depicts the sensory thresholds (pressure) before and after treatment with aspirin. In subjects without dyspeptic symptoms during aspirin treatment, sensory thresholds significantly (P 0.05) increased from to mm Hg, whereas in subjects who developed symptoms no significant increase of the sensory threshold occurred ( to mm Hg, P 0.6). Thus, differences between healthy subjects and patients with functional dyspepsia yielded significance after treatment with aspirin (P 0.02). There were no effects of treatment order (P 0.4). Wall tension. Before aspirin treatment, gastric wall tension for perception was not significantly different between healthy subjects and patients with a history of Figure 5. Study B. Thresholds for first perception after 5 days of treatment with placebo or aspirin. Medication was administered in a randomized, double-blind, cross-over fashion. The leftpanel depicts functional dysplasia (FD) patients and healthy subjects without development of symptoms, the right panel indicates subjects with development of symptoms during treatment with aspirin.

6 1456 HOLTMANN ET AL. GASTROENTEROLOGY Vol. 123, No. 5 functional dyspepsia (wall tension vs mm Hg/cm 2,F 0.8, P 0.8). During treatment with aspirin, wall tension was significantly lower in subjects with symptoms compared with subjects without symptoms (P 0.05). Combined post hoc data analysis of sensory data of studies A and B. Twelve of 14 patients and 4 of 14 healthy subjects developed symptoms during aspirin treatment ( 2 9.1, P patients vs. healthy subjects). The combined analysis of studies A and B confirmed an association between alterations of sensory thresholds and the development of symptoms on aspirin treatment. In subjects who developed symptoms during aspirin treatment, no significant changes of sensory thresholds occurred (P 0.15), whereas in subjects who remained asymptomatic, sensory thresholds significantly increased compared with sensory thresholds measured before aspirin or after placebo treatment (P 0.001). No significant differences between studies A and B were detected (P 0.5). Discussion Treatment with aspirin and traditional NSAIDs usually causes acute mucosal lesions. 1 Only a proportion of patients with mucosal lesions develop dyspeptic symptoms. The mechanisms involved in the development of dyspeptic symptoms whether or not there are mucosal lesions remain unknown. Similarly, the mechanisms that are linked to an absence of dyspepsia in patients with acute mucosal damage have not been systematically studied. The present study confirms that subjects reporting dyspepsia during treatment with aspirin do not have more severe mucosal lesions compared with subjects without symptoms. 2 However, during treatment with aspirin there appear to be distinct changes of visceral mechanosensory thresholds. Subjects not developing dyspepsia during aspirin treatment were characterized by a significant increase of sensory thresholds, whereas thresholds remained unchanged or decreased in subjects who developed symptoms. The observations suggest that the development of symptoms with aspirin may be explained in part by the lack of a normal increase of gastric mechanosensory thresholds despite the presence of mucosal erosions. On the other hand, aspirin is an analgesic, and it is well recognized that NSAID-associated gastrointestinal complications such as bleeding from NSAID ulcers may occur in the absence of any dyspeptic symptoms. 22 Thus, it might be argued that the increase of the sensory thresholds observed in this study simply reflect an analgesic effect of aspirin. However, this is unlikely because aspirin has a short plasma half-life, and the interval between the last dose of medication and the measurement of sensory thresholds was several times longer than the plasma half-life. In addition, even if a long-lasting analgesic effect of aspirin had influenced study A, this would not apply to study B in those starting on placebo. The second study, however, confirmed the initial results. The results, although novel, are consistent with other recent observations. We have reported that after repeated gastric distention of the proximal stomach, thresholds for first perception increased in healthy subjects whereas thresholds only temporarily increased in patients with functional dyspepsia. 23 Similarly, studies in irritable bowel syndrome patients 24 have shown a decrease of sensory thresholds after repeated mechanical colorectal stimulation (referred to as rectal hyperalgesia). These data, therefore, suggest that there may be a failure of visceral sensory thresholds to appropriately respond to a noxious stimulus in a subset of patients with functional gastrointestinal disease. In the present study, the development of symptoms and a failure of sensory thresholds to increase were most evident in the patients with functional dyspepsia; hence, in the initial single-blind trial (study A), 6 of 8 patients with functional dyspepsia developed symptoms during treatment with aspirin, and only one patient showed an appreciable increase of their sensory thresholds. In contrast, 6 of 8 healthy volunteers had a clear increase of gastric sensory thresholds. These findings were confirmed in the randomized, blinded, cross-over trial. Interestingly, the changes of sensory thresholds were not prolonged; in subjects with a follow-up measurement off aspirin, sensory threshold values returned to the baseline levels within a week. Thus, we speculate that after induction of mucosal lesions, stimulation of visceral afferents is increased (possibly via direct exposure of gastric acid to free mucosal nerve endings). If endogenous pain control systems are activated (as indicated by an increase of mechanosensory thresholds), these lesions do not result in dyspeptic symptoms. However, if there is a lack of an increase of visceral sensory thresholds (i.e., if sensory thresholds remain unchanged), the risk that dyspeptic symptoms will occur may be increased. In the present study we assessed sensory thresholds and did not measure pain thresholds. It thus might be argued that the measurement of pain thresholds would have yielded different results. However, a study design incorporating measurements of pain thresholds, exposing the subjects repeatedly to pain, would have imposed the risk for dropouts and hence selection bias. On the other hand, abdominal discomfort is an important symptom induced by aspirin, as we observed in this study. Thus,

7 November 2002 DYSPEPSIA DEVELOPMENT WITH ASPIRIN 1457 we focused on thresholds for discomfort rather than measuring thresholds for pain. Moreover, recent data have shown a close correlation of thresholds for first perception and pain 25 and thus pain and discomfort thresholds are most likely closely correlated. The sensory thresholds in patients with functional dyspepsia and healthy subjects were not significantly different at entry. Because all patients with functional dyspepsia were asymptomatic at entry, we conclude that the similarity of sensory thresholds at baseline was probably attributed to a lack of disease activity in the patient population. Our observations do not contradict the now widely accepted concept that sensory dysfunction is important for the development of symptoms in patients with functional gastrointestinal disorders. 26 Indeed, when symptoms occurred in patients while taking aspirin there were clear differences in terms of sensory thresholds between subjects with and without symptoms. It might be argued that patients with functional gastrointestinal disorders and fluctuating symptoms represent a very specific subgroup of patients. However, the majority of patients with functional dyspepsia have fluctuating symptoms. 27 Thus, we feel confident that our study population is representative of at least a reasonable proportion of patients with functional dyspepsia. In study A, the order of treatment was fixed and only half of the patients and healthy subjects had a placebo follow-up evaluation. Thus, it might be argued that the sequence order influenced our findings. However, at follow-up evaluation on placebo, the thresholds unequivocally returned to normal (P 0.4 vs. baseline); although it is noteworthy that 3 of 4 healthy subjects had numerically higher values compared with baseline before aspirin, the differences were not significant. In study A, the distinct effects of aspirin and the normalization of thresholds after placebo treatment suggest that bias was not of importance. Furthermore, our results were confirmed in the randomized, cross-over trial (study B) that we conducted to further validate our findings. Based on our data, we speculate that normal subjects who develop symptoms on treatment with aspirin have an impaired ability to increase their gastric sensory thresholds, and this is a feature they share in common with patients with functional gastrointestinal disorders. A disturbance of endogenous sensory modulation systems is a potential explanation for our findings. Based on previous animal studies, it appears most likely that mechanical stimulation activates descending corticospinal, bulbospinal, and spinal inhibitory systems. 11 These mechanisms probably protect against the development of sensitization or even increase visceral sensory thresholds, but may be disturbed in functional dyspepsia. Afferents activating the putative descending inhibitory pathways are most likely mediated by the vagal nerve 28 and vagal dysfunction is closely linked with visceral hyperalgesia in functional dyspepsia. 7,29 In healthy asymptomatic subjects, activation of descending inhibitory pathways may prevent sensitization (resulting in lowered sensory thresholds) whereas failure to activate these descending pathways may be responsible for the development of symptoms in patients with functional dyspepsia. In addition, inflammatory mediators that are released during aspirin treatment also may influence the outcome. Although very little is known about the potential underlying mechanisms that could explain the sensory changes, it might be speculated that there are differences in the activation of antinociceptive pathways, and, in particular, opioidergic pathways. This hypothesis currently is being explored. We further speculate that the treatment of some patients with functional dyspepsia may require normalization of the capability to respond appropriately to a sensitizing stimulus rather than simply an increase of sensory thresholds. This may account in part for the relatively disappointing results with some visceral analgesics such as opioid agonists in functional dyspepsia. 30 Because aspirin alters prostaglandin metabolism and prostaglandins appear to be involved in the regulation of gastrointestinal motility (e.g., prostaglandins administered orally can induce phase III complexes 16 ), any effects of aspirin on symptoms could be mediated via alterations of motor function. Indeed, aspirin treatment temporarily but significantly decreased the gastric half-emptying time. However, the changes of gastric emptying were very modest and not associated with the development of symptoms; indeed, gastric emptying was still in the normal range. In summary, subjects on aspirin remaining free of symptoms appear to be characterized by an increase in gastric sensory thresholds, whereas subjects developing symptoms may fail to significantly change gastric sensory thresholds. The data provide a potential explanation for why some subjects on aspirin report symptoms whereas others remain totally asymptomatic. Although the lack of a significant increase in sensory thresholds appears to be more likely in subjects with functional dyspepsia, we know no other studies that have assessed whether patients with functional dyspepsia are more likely to develop acute dyspepsia than healthy subjects when starting aspirin, but if so the mechanisms deserve further exploration.

8 1458 HOLTMANN ET AL. GASTROENTEROLOGY Vol. 123, No. 5 References 1. Jones RH, Tait CL. Gastrointestinal side effects of NSAIDs: a community-based study. Gut 1993;34:S Collins AJ, Davies J, Dixon SA. Contrasting presentation and findings between patients with rheumatic complaints taking nonsteroidal anti-inflammatory drugs and a general population referred for endoscopy. Br J Rheumatol 1986;25: Holtmann G, Goebell H, Talley NJ. Dyspepsia in consulters and non-consulters: prevalence, health-care seeking behaviour and risk factors. Eur J Gastroenterol Hepatol 1994;6: Steering Committee of the Physicians Health Study Group. Final report on the aspirin component of the ongoing physicians health study. N Engl J Med 1989;321: Azpiroz F, Malagelada JR. Perception and reflex relaxation of the stomach in response to gut distension. Gastroenterology 1990; 98: Lemann M, Dederding JP, Flourie B, Franchisseur C, Rambaud JC, Jian R. Abnormal perception of visceral pain in response to gastric distension in chronic idiopathic dyspepsia. The irritable stomach syndrome. Dig Dis Sci 1991;36: Holtmann G, Goebell H, Jockenhövel F, Talley NJ. Altered vagal and intestinal mechano-sensory function in chronic unexplained dyspepsia. Gut1998;42: Whitehead WE, Holtkotter B, Enck P, Hoelzl R, Holmes KD, Anthony J, Shasbin HS, Schuster MM. Tolerance of rectosigmoid distension in irritable bowel syndrome. Gastroenterology 1990; 90: Bradette M, Pare P, Douville P, Morin A. Visceral perception in health and functional dyspepsia. Crossover study of gastric distension with placebo and domperidone. Dig Dis Sci 1991;36: Ritchie J. Pain from distension of the pelvic colon by inflating a balloon in the irritable bowel syndrome. Gut 1973;14: Mayer EA, GebhartGF. Basic and clinical aspects of visceral hyperalgesia. Gastroenterology 1994;107: Gebhart GF, Randich A. Vagal modulation of nociception. Am Pain Soc J 1992;1: Silverman DHS, Munakata JA, Ennes H, Mandelkern MA, Hoh CK, Mayer EA. Regional cerebral activity in normal and pathological perception of visceral pain. Gastroenterology 1997;112: Holtmann G, Gschossmann J, Neufang-Huber J, Gerken G, Talley NJ. Differences in gastric mechanosensory function after repeated ramp distensions in non-consulters with dyspepsia and healthy controls. Gut 2000;47: Stanghellini V, Tosetti C, Paternico A, Barbara G, Morselli Labate AM, Monetti N, Marengo M, Corinaldesi R. Risk indicators of delayed gastric emptying of solids in patients with functional dyspepsia. Gastroenterology 1996;110: Staumont G, Fioramonti J, Frexinos J, Bueno L. Oral prostaglandin E analogues induce intestinal migrating motor complexes after a meal in dogs. Evidence for a central mechanism. Gastroenterology 1990;98: Lanza FL, Kochman RL, Geis GS, Rack EM, Deysach LG. A double-blind, placebo-controlled, 6-day evaluation of two doses of misoprostol in gastroduodenal mucosal protection against damage from aspirin and effect on bowel habits. Am J Gastroenterol 1991;86: Ghoos YF, Maes BM, Geypens BJ, Mys G, Hiele MI, Rutgeerts PJ, Vantrappen G. Measurement of gastric emptying rate of solids by means of a carbon-labeled octanoic acid breath test. Gastroenterology 1993;104: Distrutti E, Azpiroz F, Soldevilla A, Malagelada JR. Gastric wall tension determines perception of gastric distension. Gastroenterology 1999;116: SAS Institute Inc. SAS/STAT User s guide, release Cary, NC: SAS Institute Inc., Holtmann G, Guerra G, Gschossmann J, Goebell H, Talley NJ. Perception of gastric distension: influence of the mode of distension on perception thresholds. Dig Dis Sci 1996;40: Tenenbaum J. Non-steroidal anti-inflammatory drugs (NSAIDs) cause gastrointestinal intolerance and major bleeding or do they? Clin Invest Med 1987;10: Holtmann G, Gschossmann J, Neufang-Huber J, Gerken G, Talley NJ. Differences in gastric mechanosensory function after repeated ramp distensions in non-consulters with dyspepsia and healthy controls. Gut 2000;47: Munakata J, Naliboff B, Harraf F, Kodner F, Lembo T, Chang L, Silverman DH, Mayer EA. Repetitive sigmoid stimulation induces rectal hyperalgesia in patients with irritable bowel syndrome. Gastroenterology 1997;112: Holtmann G, Goebell H, Talley NJ. Impaired small intestinal peristaltic reflexes and sensory thresholds are independent functional disturbances in patients with chronic unexplained dyspepsia. Am J Gastroenterol 1996;91: Hotz J, Enck P, Goebell H, Heymann-Monnikes I, Holtmann G, Layer P. Consensus report: irritable bowel syndrome definition, differential diagnosis, pathophysiology and therapeutic possibilities. Consensus of the German Society of Digestive and Metabolic Diseases. Z Gastroenterol 1999;37: Talley NJ, Weaver AL, Zinsmeister AR, Melton LJ III. Onset and disappearance of gastrointestinal symptoms and functional gastrointestinal disorders. Am J Epidemiol 1992;136: Randich A, Ren K, Gebhart GF. Electrical stimulation of cervical vagal afferents. II. Central relays for behavioral antinociception and arterial blood pressure decreases. J Neurophysiol 1990;64: Greydanus MP, Vassallo M, Camilleri M, Nelson DK, Hanson RB, Thomforde GM. Neurohormonal factors in functional dyspepsia: insights on pathophysiological mechanisms. Gastroenterology 1991;100: Read NW, Abitbol JL, Bardhan KD, Whorwell PJ, Fraitag B. Efficacy and safety of the peripheral kappa agonist fedotozine versus placebo in the treatment of functional dyspepsia. Gut 1997;41: Received July 2, Accepted August 8, Address requests for reprints to: Gerald Holtmann, M.D., Professor of Medicine, Department of Internal Medicine, Division of Gastroenterology and Hepatology, University of Essen, Hufelandstrasse 55, Essen, Germany. g.holtmann@uni-essen.de; fax: (49) Supported by a grant from Deutsche Forschungsgemeinschaft, grant number Ho 1193/3-7. The authors thank J. Neufang-Hüber for her skillful assistance, and A. Kaune is greatly appreciated.